Wouldn't it be nice... (the LazyWhale algorithm)

[oda.krell]

LW Live Signals:

(don't take them too seriously, I'd suggest)

Buy @239-241, 2015-05-03

Sell @236-244, 2015-04-09

Buy @236, 2015-02-13

Sell @311, 2014-12-19

Buy @372, 2014-11-10

[1715684187]

1715684187

[1715684187]

1715684187

[1715684187]

1715684187

[Livermore]

What is the basis of your algorithm?

[oda.krell]

What is the basis of your algorithm?

Already answered that, as far as I will answer it for now...

Q & A

Q: Is this a momentum-following strategy?

A: Yes and no. At the core of the strategy, there are several momentum signals, but they are combined in a not completely obvious way, I would claim. So the complete algorithm doesn't have much in common with, say, a classical exponential moving average strategy.

[...]

Q: Are you going to tell us the exact parameters of the method?

A: Not right now. Think of this as a 'public test' of the method, to see how it performs. If it fails, you're free to ridicule me publicly as well

[Raystonn]

Oda, would you mind sharing the tools you are using for automating this?  I'm looking to recreate your workflow.  I use Tradestation for equities, options, and futures.  But I haven't found anything decent for Bitcoin yet.

[oda.krell]

Oda, would you mind sharing the tools you are using for automating this?  I'm looking to recreate your workflow.  I use Tradestation for equities, options, and futures.  But I haven't found anything decent for Bitcoin yet.

I am subscribed to Tradingview myself, which is an okay site, but I started noticing that most of my "idol" traders are using Sierrachart. D'oh! I believe that the ability to code your own strategies is at least as developed as on TV.

Other than that, some of what I call my 'tests' are ad hoc written programs (in Java, usually) that I crudely hack together to test some particular idea on some data set. I'm almost sure there are more efficient ways to go about this than what I do (R comes to mind, but I'm not good enough at it), so I don't recommend that approach.

[sgbett]

nice tread!

Although I am buy and hold, risk management means that during every run up I took money off the table. If I had the foresight to have just held everything since when I started I would be better of technically (by a factor of about x 3) but in selling on the way up and buying on the way down I've ended up in a pretty nice spot, have paid some bills and have a couple of bitcoin for a rainy day. If I had the discipline to have followed this algo...

...hindsight's a tough old bitch though eh

[sgbett]

Speaking of hindsight, my BFL SC60 just found its *second* block   alas I'm in a pool...

oh, how I laughed!

[oda.krell]

Speaking of hindsight, my BFL SC60 just found its *second* block  ::) alas I'm in a pool...

oh, how I laughed!

You're 'legendary', in it since 2011, and selling bits and pieces only brought you down by a factor of 3 from that 2011 entry point...

I'm going to assume you're doing well enough, pool or not pool :P

[vuduchyld]

Sweet!  I hope the algo produces another buy signal soon, because I'm as in as I need to be right now.

May follow the next sell signal, but only with a part of my stash...and probably only if I decide to buy more in the near future.

[sgbett]

Speaking of hindsight, my BFL SC60 just found its *second* block  alas I'm in a pool...

oh, how I laughed!

You're 'legendary', in it since 2011, and selling bits and pieces only brought you down by a factor of 3 from that 2011 entry point...

I'm going to assume you're doing well enough, pool or not pool

Heh, I'm doing OK.

On the one hand I could lament that I barely had the cash to feed myself in 2011, so I really was scraping around for something *anything* to buy a few BTC with!

On the other hand it was only a dollar a coin

I know it sounds like I'm just playing it down but I've made for more money actually working than I have with BTC (...for now at least!)
 
Hindsight would have probably starved me to death!

[sgbett]

Trying to get into mining has probably been the worst decision I made throughout it all. At the time it made sense. There's that hindsight again

[sgbett]

...hold on a sec... Legendary!? When did that happen! 

They must let anyone in thesedays. All I ever did was trash talk spec forum. Legendary... lol.

[oda.krell]

Trying to get into mining has probably been the worst decision I made throughout it all. At the time it made sense. There's that hindsight again

It was purely me being too lazy to set up a mining rig that prevented me from pre-ordering a BF lab or joining one of the ASIC batch orders last year.

Can you imagine how much I reward my laziness these days for making that call?!

EDIT: what are *you* drinking, btw? JW BL for me

[sgbett]

Not drinking tonight. Just up late trying to work but struggling. Procrastination is a killer. I might have a camomile tea I think.

When I first heard about BTC they were about $0.30 and you could (just) CPU mine. I started to set it up but got distracted. I think the fact I missed that mining opportunity might have made want to "put it right" hehe. I had all sorts of plans of running solar arrays in spain to power a mining farm. Mind blowing how fast it all moved. Surely some indication of what is to come if market penetration of the product follows the same pattern...?

Anyway that little black box that cost me 25BTC is now straining out 0.6mBTC a day, which makes me want to run solo to see if I get lucky but TGB says its a 100 years for me to find a block!?

Watch me hit a third before the year is out!

[Wary]

You've got your own monkey! Congrats! Have you given a name to him?

[RyNinDaCleM]

Nice find, oda!

Is this something that can be scaled down to a shorter time frame, if only to have a larger testing sample for sustainability? I'm curious to see how this works out. Watching closely!

[oda.krell]

You've got your own monkey! Congrats! Have you given a name to him?

Hehe, the other monkey had a bit too many 'buy' signals during the last months right in the middle of a downtrend, though it looks like his last call before 275 might come true. I suspect he's using more of a mean reversion way of thinking, which would make his monkey more aggressive than what I'm trying here, which is inherently conservative.

A name?

How about LazyWhale™?

[oda.krell]

Nice find, oda!

Is this something that can be scaled down to a shorter time frame, if only to have a larger testing sample for sustainability? I'm curious to see how this works out. Watching closely!

Thanks Ryn. Means something coming from you. Also, excellent question, cutting right to the core:

Yes, the low number of signals also means it's harder to rule out that the results were just produced by overfitted parameters.

I'd argue that, watching where the signals fall in particular, it is unlikely that this is the case, as the 'sell' reliably comes in around the 1st half or so of any major trend breakdown, and the 'buy' reliable comes early into each lasting recovery, but that's just my "qualitative" judgement, not a quantifiable result I admit.

As for a higher frequency version: It should be easy enough to increase frequency for the momentum portion of the method, just a search for parameters with different constraints. The other part of the method is more difficult to adjust for a higher signal frequency, I would have to think about it a bit. ... As I write this, I have an idea maybe how to go about it. Will report back when I have a test for the adjusted method.

[Wexlike]

I really love this strong hodl approach while at the same time taking advantage of the bear market. Good job Oda!

I am curious, could this swing trading also work in an altcoin market?

Would be interesting if you could backtest your algorithm for example in the litecoin market.

Again, good job and a great read!

[Brmaster]

Make money like doing it? I doubt